An Optimized Business Service Directory for the ESB Platform in SOA

Abstract

Service Oriented Architecture (SOA) helps toprovide value added services to the user with agility by following a publish-find-compose technique. Business Service Directory is one of the key components of Enterprise Service Bus and its implementation involves many challenges. One of the essential challenges is how to provide a suitable set of service candidates faster. This paper proposes a business service directory which employs ranking based, quality driven publishing and searching. The proposed business service directory is a fuzzy expert system which requires no human intervention. The implementation of optimized business service directory is composed of two steps, first the rule base is designed which contains auto generated rules to rank the web service. Fuzzy clustering and PSO is used to not only generate rules automatically using training dataset but also removing redundant and less effective rules. The next step is to design inference engine which triggers the rules in the rule base. The proposed Business service directory minimizes administrative overhead and increases usability by locating the best web service among the large number of functionally equivalent web services and it’s also adaptive in nature.